Mechanistic Investigation of the Nickel-Catalyzed Carbonylation of Alcohols

Sara Sabater, Maximilian Menche, Tamal Ghosh, Saskia Krieg, Katharina S. L. Ru

Article

Mechanistic Investigation of the Nickel-Catalyzed Carbonylation of

Alcohols

ck, Rocco Paciello,

Ansgar Schafer, Peter Comba, A. Stephen K. Hashmi, and Thomas Schaub*

Cite This: https://dx.doi.org/10.1021/acs.organomet.0c00082

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: The carbonylation of alcohols represents a

straightforward and atom-eﬃcient methodology for the prepara-

tion of carboxylic acids. It is desirable to perform these reactions

under precious metal-free and low-pressure conditions, with

regioselectivity control. In this work, we present a detailed

mechanistic study of a catalytic system based on NiI , which can

carbonylate benzylic alcohols in a highly regioselective manner to

the corresponding branched carboxylic acids, core motifs for

nonsteroidal drugs. The combination of catalytic amounts of nickel and iodide is crucial for eﬃcient catalytic and regioselective

conversion. Quantum-chemical computations were used to evaluate the underlying mechanistic processes. They revealed that a

combination of two mechanisms is responsible for the observed reactivity and that the oxidative addition of alkyl halides to the

Ni(0) species follows a radical oxidation pathway via two one-electron steps.

INTRODUCTION

is necessary. Industrially, this is achieved using a Pd-PPh₃

catalyst and running the process in HCl (20%) at 100 bar of

CO to achieve the desired branch selectivity, necessitating the

use of reactors made from expensive corrosion-resistant alloys to

withstand the harsh conditions. Driven by the increased cost of

palladium, there is a need for the development of alternative

routes, ideally using cheaper earth-abundant metals, less

corrosive systems, and CO pressures below 100 bar.

■

Catalytic carbonylation reactions of oleﬁns and alcohols are of

signiﬁcant interest for both academia and industry, as they

provide a straightforward methodology for accessing carboxylic

acids. At pressures below 100 bar, precious metals such as

−6

7,8

9,10

palladium,

rhodium,

or iridium

are usually used

industrially in the production of bulk chemicals such as acetic

acid or methylpropionate. With nonprecious metals, harsher

conditions are normally necessary. For example, the BASF

process for propionic acid production via the carbonylation of

In a seminal report, Rizkalla studied the carbonylation of

methanol to acetic acid using a Ni-PPh catalytic system in

which milder conditions are required in the presence of catalytic

ethylene, where Ni(CO) is used as the catalyst, requires a CO

promoters.

pressure of 300 bar and temperatures of 250−320 °C. These

conditions are, in most cases, not compatible with more complex

substrates. However, carbonylations are also of high relevance

for the production of active pharmaceutical ingredients (APIs),

such as ibuprofen via carbonylation of the corresponding

alcohol, 1-(4-isobutylphenyl)ethanol (Scheme 1). In this

reaction, a high selectivity toward the branched carboxylic acid

Despite some other reports on the carbonylation of

15,16

17,18

methanol,

ethanol, and n-propanol,

the use of nickel

for the preparation of more complex carboxylic acids, such as 2-

arylpropionic acids, has not been extensively studied, and it has

19,20

only been reported in two patents.

Therefore, our objective

was to perform a detailed mechanistic study, aimed at a better

understanding of the system in order to evaluate the potential

applications and limitations for industrial or academic purposes.

Here, we present our investigation on the direct carbonylation of

alcohols using a simple, less-studied catalytic system based on

Scheme 1. Synthetic Route for the Industrial Production of

Ibuprofen

21,22

NiI and the inexpensive tri-n-butylphosphine (TBP) ligand

Received: February 6, 2020

XXXX American Chemical Society

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Table 1. Evaluation of Catalytic Components

Entry

[Ni] (4 mol %)

NiI₂

[L] (8 mol %)

Additive

Yield 1a (%)

Yield 1b (%)

Yield 1c (%)

TBP

TBPHI

Ni(COD)₂

TBP

LiI

LiI + HI

NiI (2 mol %)

TBP (4 mol %)

LiI

A mixture of 1-phenylethanol (8.3 mmol), [Ni] (2 or 4 mol %), TBP (4 or 8 mol %), and the corresponding additive (8 mol %) when indicated

was pressurized with CO at 50 bar and stirred at 120 °C for 20 h. Yields were determined by GC-FID chromatography using anisole as an internal

standard after derivatization with N-methyl-N-(trimethylsilyl)triﬂuoroacetamide. 57% in water.

under relatively mild reaction conditions. Remarkably, the

system is selective toward the branched product when benzylic

alcohols are used as substrates and can generate a range of

carboxylic acids starting from readily accessible alcohols.

of HI and 1 equiv of CO₂. The results in Table 1 show that

Ni(COD)₂is able to perform the carbonylation of 1-

phenylethanol when an external source of HI is present in the

reaction medium (entries 2, 3, and 7), suggesting that the

reduction of Ni(II) to Ni(0) is the ﬁrst step in the catalytic cycle.

No 2-phenylpropionic acid was obtained in the absence of either

TBP or HI (entries 4−6), indicating that the three-component

system Ni/L/HI is needed together in the catalytic cycle. The

activity signiﬁcantly decreased when the catalyst loading was

decreased to 2 mol % (entry 8). In this case, 1a was obtained in

only 16% yield, whereas 1b and bis(1-phenylethyl)ether were

the main products. The ether might originate from a reaction of

RESULTS AND DISCUSSION

■

Inspired by the work patented by Mitsubishi on the carbon-

ylation of 1-(4-isobutylphenyl)ethanol to ibuprofen using

NiI₂, our investigation began by studying the direct carbon-

ylation of alcohols. 1-Phenylethanol (1) was chosen as a

benchmark substrate for the optimization. In a ﬁrst catalyst

screening, diﬀerent Ni sources were evaluated. The experiments

were carried out using 4 mol % of Ni, 8 mol % of TBP, 50 bar of

CO at 120 °C for 20 h, under solvent-free conditions. Self-made

stainless steel autoclaves from Swagelok parts ﬁtted with a

frequently replaced Teﬂon insert were used in order to avoid any

cross-contamination (see Supporting Information). Diﬀerent

halogen-containing nickel precursors can perform the carbon-

ylation of 1 to the corresponding 2-phenylpropionic acid (1a).

To our delight, the carbonylation was regioselective, aﬀording

-phenylethanol with 1-iodophenylethane, formed with HI in

situ generated under the reaction conditions. Aiming at an

enhancement of the catalytic activity, we performed a screening

of additives. As known from the carbonylation of methanol,

ethanol, and n-propanol, LiI can have a positive eﬀect on the

reaction outcome. 1a was obtained in 70% yield using a catalyst

loading of 2 mol % of NiI , 4 mol % of TBP, and 8 mol % of LiI

entry 9). Other additives that have been claimed to be

beneﬁcial for the carbonylation of methanol and ethanol did

not enhance the activity in our case (see Supporting

Information ).

Role of HI. The formation of iodoethylbenzene from 1-

phenylethanol and HI has already been reported. In order to

investigate the role of HI, the carbonylation of cyclohexanol

using Ni(COD) and cyclohexyl iodide as the iodide source and

promoter was carried out (Scheme 2). The reaction outcome

(

the branched product 1a in 83% yield when NiI was used. Full

conversion of 1-phenylethanol was achieved and 1,3-diphenyl-

but-1-ene (1b) and ethylbenzene (1c) were obtained as minor

byproducts in 11 and 4% yield, respectively (Table 1, entry 1).

The formation of 1b might be due to a dimerization of styrene

formed via a competitive elimination reaction of 1-phenyl-

ethanol. A detailed screening of diﬀerent Ni salts is given in the

Supporting Information. Similar to the Monsanto acetic acid

process, as well as previous reports on carbonylation using

7,24

Ni,

the halide seems to play an important role in the reaction

Scheme 2. Evaluation of the Role of HI

outcome. Therefore, it is not surprising that nickel sources such

as Ni(OAc) or Ni(OTf) , in the absence of a halide, were

inactive for the carbonylation of 1-phenylethanol. We also

carried out the reactions in the absence of nickel and observed

that all starting material was recovered. The palladium content

of the commercial NiI evaluated by inductively coupled plasma

mass spectrometry was below the detection limit of 1 ppm,

ruling out the possibility of a Pd-catalyzed reaction.

Evaluation of Catalytic Components. Nickel(0) carbon-

yls are readily formed via reductive carbonylation of nickel salts

under CO pressure, with the concomitant formation of 2 equiv

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

was similar to the one using NiI as a catalyst precursor, revealing

Table 2. Ligand Screening

that HI is needed to convert the alcohol into the corresponding

alkyl iodide, which in turn is presumed to react with the in situ

formed Ni(0) carbonyl species. Moreover, a racemic 2-

phenylpropionic acid mixture was obtained when the

enantiomerically pure (R)-1-phenylethanol was used as

substrate, as the mechanism of alkyl iodide formation follows

a S 1 pathway via an intermediate benzylic carbocation.

Ligand Screening. Methyl ethyl ketone (MEK) was the

solvent of choice for a ligand screening in our system as it is the

solvent used in industrial ibuprofen production, and it is able to

solubilize the solid ligands tested. This set of experiments was

performed using a HEL CAT7 autoclave, which allows for the

performance of seven simultaneous reactions under the same

pressure and temperature. However, it needs to be pointed out

that the results obtained are not comparable with those

performed using diﬀerent autoclaves for the experiments

above. In order to avoid any cross-contamination from previous

experiments, glass microwave vials were used and disposed after

completion of the reaction. Under the same reaction conditions,

-phenylethanol, NiI (2 mol %), MEK (8 M), 50 bar CO, and

20 °C for 20 h, the optimum Ni/TBP ratio was found to be 1:2

(

Table 2, entries 1−3). Aromatic phosphines showed an activity

higher than that of the aliphatic ones (entries 2, 4, 5, 11, and 13).

Notably, despite the optimum Ni/TBP ratio of 1:2, mono-

dentate ligands performed much better than bidentate ligands

(

compare entries 4 and 5 with 6−10). From the series of

bidentate ligands tested, the activity was slightly increased with

the backbone length (entries 7−10). Our density functional

theory (DFT) calculations (vide infra) show that the transition

states with only one phosphine ligand coordinated to the nickel

center are lower in energy than the corresponding transition

states with two phosphines, thus we further investigated

monodentate phosphine ligands. Interestingly, when DavePhos,

which contains an amine functionality in its skeleton, was tested

as the ligand, the same catalytic activity was observed when

either 1 or 2 equiv was used (entries 11 and 12). This suggests

that only one phosphine is coordinated to the nickel center,

whereas the other could be acting as a base. To conﬁrm this

observation, we performed the same ratio screening using

CyJohnPhos, analogous to DavePhos but without functional

groups attached to the skeleton, and 2-(dicyclohexylphosphi-

no)-2′,4′,6′-triisopropylbiphenyl (XPhos). The catalytic activity

was dramatically decreased when only 1 equiv of phosphine was

used (compare entries 14, 15 and 18, 19). In contrast, the

catalytic activity was maintained when an equimolar ratio of

A mixture of 1-phenylethanol (8.3 mmol), NiI (2 mol %), and the

corresponding ligand (2 or 4 mol %) in MEK (8 M) was pressurized

with CO at 50 bar and stirred at 120 °C for 20 h. Yields were

determined by GC-FID chromatography using anisole as an internal

standard after derivatization with N-methyl-N-(trimethylsilyl)-

triﬂuoroacetamide. 2 mol % of Hu

of 2,6-lutidine was added.

phosphine and the noncoordinating Hu

was used (entries 16 and 17). The reaction in the absence of

phosphine and 2 equiv of Hunig’s base instead resulted in an

nig’s base or 2,6-lutidine

nig’s base was added. 2 mol %

inactive system. These results further support our hypothesis

that one phosphine is coordinated to the metal center in the

catalytic cycle while the other acts as a base.

−

The use of mono- and bidentate N-donor ligands, as well as

phosphites, resulted in inactive systems (see Supporting

Information) in contrast with what was previously reported

halogenotricarbonyl anion [Ni(CO) I] Li, and that the

complex reacts in the presence of phosphine to generate nickel

carbonyl phosphine complexes (Scheme 3). This Ni(0)

halogenotricarbonyl anion has already been suggested as an

active species for the carbonylation of methanol and ethanol

to acetic and propionic acid and aromatic halides to the

for the carbonylation of ethanol.

Optimum LiI Concentration. The optimum LiI concen-

tration was evaluated using 1 mol % of NiI and 2 mol % of TBP

in MEK (8 M). Results in Table 3 show that higher

concentrations of LiI enhance the catalytic activity. When the

carbonylation was performed in the absence of phosphine, 1-

phenylethanol was still carbonylated to 2-phenylpropionic acid.

corresponding carboxylic acids.

Our DFT studies (vide infra) show that the energetic barriers

for the global reaction are similar starting from either a Ni(0)

halogenotricarbonyl anion or a Ni(0) carbonyl phosphine

species. In our system, at low concentrations of LiI, the

It is known that Ni(CO) reacts with LiI to the Ni(0)

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Table 3. Variation of LiI Concentration

[

LiI]

Yield 1a (%)

Yield 1a (%) (0 mol % of TBP; 1 mol % of

Entry

(x mol %)

(2 mol % of TBP)

(0 mol % of TBP)

(1 mol % of TBP)

nig’s base)

A mixture of 1-phenylethanol (8.3 mmol), NiI (1 mol %), TBP (0−2 mol %), Hunig’s base (1 mol %) when indicated, and LiI (x mol %) was

pressurized with CO at 50 bar and stirred at 120 °C for 20 h. Yields were determined by GC-FID chromatography using anisole as an internal

standard after derivatization with N-methyl-N-(trimethylsilyl)triﬂuoroacetamide.

Scheme 3. Formation of Halogenocarbonylnickel Species

radical process. More recently, the oxidative addition of alkyl

halides to Ni(0) phosphine complexes has been studied in detail

−

[

Ni(CO) I] Li

31−38

for cross-coupling and reductive carbonylation reactions.

The reaction appears to follow a stepwise process, via two one-

electron steps, involving alkyl radicals and halonickel(I)

metalloradical species. Radical clocks such as iodomethylcyclo-

propane have been used to support such types of mecha-

35,39

nisms.

In the presence of radicals, an opening of the

cyclopropane ring should be observed. In our case, due to the

increased temperatures of 200 °C required for the carbonylation

of aliphatic alcohols (vide infra), the ring opening occurred even

in the absence of catalyst, inducing misleading results.

Nevertheless, the presence of Ni(I) species was conﬁrmed by

electron paramagnetic resonance (EPR) spectroscopy of a

reactive mixture of Ni(PPh ) (CO) and iodocyclohexane. The

combination of LiI and 2 equiv of phosphine performed better

than only LiI (columns 1 and 2, entries 1−3). The experimental

and computational data suggest that two diﬀerent mechanisms

can work in concert when phosphine and LiI are both present in

the reaction medium. When the reactions were carried out with

equiv of phosphine or 1 equiv of Hunig’s base, the outcome

was similar (compare columns 3 and 4). This result suggests that

the role of the second equivalent of phosphine is to neutralize 1

equiv of HI, generated in the reduction step, forming a

EPR spectrum (Figure 1 ) of the reaction mixture was run at 203

quaternary phosphonium salt that is able to stabilize the ionic

−

species [Ni(CO) I] . At higher concentrations of LiI, the

activities are quite similar, indicating that the phosphine and the

base are playing a negligible role in the global outcome (entries 4

and 5). This observation was conﬁrmed by performing a ligand

screening using 20 mol % of LiI (see Supporting Information).

The results indicate that there is practically no diﬀerence in

activity with some of the phosphines tested, suggesting that, in

that case, only the “anionic mechanism” is acting on the overall

process and the phosphine has no real inﬂuence. When the

chelating 2-carbon backbone phosphines dCype and dppe were

used, the activity decreased by 50%. In addition, mono- and

bidentate N-donor ligands also resulted in inactivity. This sharp

decrease of activity is possibly caused by a strong coordination of

the ligands to the Ni center that inhibits the formation of the

active Ni(0) halogenotricarbonyl anion.

Figure 1. X-band (9.638417 GHz) EPR spectrum at 203 K of the

reaction mixture of Ni(PPh ) (CO) and iodocyclohexane in methyl

Di ﬀ erent iodide salts were also tested at 20 mol % loading (see

Supporting Information ). As expected, the reaction does not

ethyl ketone at 90 °C.

take place in the absence of either LiI or NiI . Sodium and

tetrabutylphosphonium iodide aﬀorded 1a in 66 and 39% yield,

respectively, whereas ammonium iodide salts were inactive. The

reason for inhibition by nitrogen-containing salts is unknown at

this stage, but it is in agreement with the inactivity shown by N-

donor ligands.

Mechanistic Investigation. In the carbonylation of benzyl

halides, Stille suggested that the oxidative addition of benzyl

halides to Ni(0) phosphine complexes might proceed via a

K and shows a resonance at g_e_f_f= 2.15, consistent with the

presence of Ni(I)-phosphine species (for a simulation of the

33,40

spectrum, see the Supporting Information).

Furthermore,

the EPR spectra of the reaction educts did not exhibit any EPR

signals. One should note that commonly EPR-silent Ni(II)

complexes with broad signals have also been reported

previously.

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

We also performed the reaction in the presence of a range of

radical scavengers or radical chain initiators (see Supporting

Information for details). Typical radical chain initiators such as

AIBN or ascorbic acid did not have any eﬀect on the reaction

outcome, ruling out the possibility of a radical chain reaction

mechanism. Typical radical scavengers such as TEMPO or

Scheme 4. Catalytic Cycle for the Carbonylation of R−I

(Iodoethylbenzene) via “Stepwise” Oxidative Addition by the

Ni(I) Complex: ΔG in kJ mol , RI-PBE0-D3(BJ)/def2-

QZVPP//BP86/def2-SV(P), COSMO-RS (1-

Phenylethanol)

393

−1

Bu SnH reacted with the precatalyst mixture, resulting in

inactive systems. Hydroquinone reacted with the substrate and

HI to yield the Friedel−Crafts alkylation product. All other

additives tested resulted in a slight decrease of activity. These

results are in agreement with the reported work of Kochi on the

mechanism of oxidative addition of aromatic halides to Ni(0)

complexes. In view of our results, we believe that the generated

phenylethyl radical preferentially reacts with the Ni(I) complex

without ever escaping the solvent cage. Therefore, the generated

radicals are quickly consumed before any formation of radical-

based side products or any deactivation by radical scavengers

can take place.

DFT Studies on the Reaction Mechanism. To further

understand the underlying processes, a detailed investigation of

the reaction mechanism was carried out on the basis of density

functional theory (see Supporting Information for details).

Mechanistic pathways via two one-electron steps have been

computationally evaluated in previous publications, mainly

focusing on a variety of alkyl−alkyl coupling reac-

5,36,39,42−50

tions.

Comparable mechanisms have been proposed

for the carbonylation of alkyl iodides with palladium, where the

single-electron transfer (SET) is initiated photocatalytically.

The subsequent reaction steps are very similar to those proposed

−1

(

ΔG = 114.0 kJ mol ), which has also been computationally

51,52

42,43 53

for nickel and considered for the system presented here.

The

studied by other groups for Ni

assumed to undergo barrierless homolytic cleavage of the R−I

bond to yield the Ni(I) complex C and the corresponding

phenylethyl radical (ΔG = 73.7 kJ mol ). It should be

mentioned that exchange of iodide on the alkyl group in a S 2-

type fashion can be excluded due to a lack of suitable orbitals on

the Ni fragment. Instantaneous recombination of R and the

and Pd complexes. B is

involvement of alkyl radicals generated by SET from the

corresponding alkyl halides in Ni-catalyzed reactions has been

39,45−49

−1

suggested by diﬀerent research groups.

However, to the

best of our knowledge, an in-depth comparison between

oxidative addition/reductive elimination and potential radical

pathways has not been carried out yet. Additionally, the aim of

our investigation was to evaluate the role of the ligand and LiI in

order to identify the ideal ligands and additives for these

reactions.

At the beginning of our computational studies, the stability of

potential key intermediates was investigated. The involvement

of two-coordinate intermediates could quickly be ruled out.

Formation of three-coordinate species is also substantially

endergonic but is within a feasible energy range for the presented

reaction conditions. The Gibbs free energies of four-coordinate

species are similarly endergonic. However, their formation via

•

Ni(I) complex C leads to the penta-coordinate complex D

−1

(ΔG = 111.1 kJ mol ), which can undergo CO insertion

⧧

−1

(TS-E; ΔG = 131.7 kJ mol ) to yield a signiﬁcantly more

stable tetra-coordinate acyl nickel complex F(ΔG = 69.7 kJ

−1

mol ). Other potential recombination pathways were consid-

ered but were found to be higher in energy (see Supporting

−1

Information ). CO addition forming G (ΔG = 75.6 kJ mol )

⧧

−1

and reductive elimination via TS-H (ΔG = 106.8 kJ mol ) lead

−1

to the R(CO)I adduct I (ΔG = 93.0 kJ mol ), which quickly

liberates the product and adds TBP to regenerate complex A

−1

“classical” oxidative addition would require the previously

(ΔG = −17.2 kJ mol ). A “classical” two-electron oxidative

addition with a subsequent CO insertion step was also

investigated but found to be signiﬁcantly disfavored (TS-J;

excluded presence of two-coordinate species (see Supporting

Information ).

⧧

−1

As it was not possible to determine the exact coordination

environment experimentally (equivalents of phosphine-coordi-

nated vs equivalents of CO-coordinated), the mechanistic

pathways were calculated with zero, one, and two molecules of

TBP coordinated to the Ni center. The activation barrier was

found to be lowest when one phosphine is coordinated to the

complex, suggesting that 1 equiv of TBP is suﬃcient to drive the

reaction (see Scheme 4), which also is supported by experiments

ΔG = 212.4 kJ mol ; Scheme 5, blue pathway). From RI

⧧

adduct B, oxidative addition will take place via TS-J (ΔG =

−

212.4 kJ mol ), forming the penta-coordinate Ni(II) complex

−1

D (ΔG = 111.1 kJ mol ). CO insertion (and all subsequent

steps) will take place as for the one-electron step pathway (TS-E;

⧧

−1

ΔG = 131.7 kJ mol ) and leads to the signiﬁcantly more stable

−1

tetra-coordinate complex F (ΔG = 69.7 kJ mol ). Similarly

high two-electron oxidative addition barriers have been reported

by Bottoni et al., which investigated the formation of β,γ-

using 1 equiv of phosphine and Hunig’s base (vide supra). Both

42,43

the activation barriers for equivalent mechanistic scenarios

based on systems with zero (three CO ligands) or two TBP

ligands are higher in energy (see Schemes S3 −S6 in Supporting

Information). The reaction proceeds from Ni(CO) (TBP) (A)

unsaturated acyl derivatives via a Ni(0)/Ni(II) pathway.

These considerations were also transferred to the depe ligand

(1,2-bis(diethylphosphino)ethane), which was used as a model

for the barely active dppe ligand (1,2-bis(diphenylphosphino)-

ethane). The calculated activation barrier of the rate-

via phosphine dissociation and formation of the R−I adduct B

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Scheme 5. Potential Energy Surface for the Carbonylation of R−I (Iodoethylbenzene)

393

−1

The energetically unfavored two-electron oxidative addition/CO insertion pathway is shown in blue for comparison; ΔG in kJ mol ; RI-PBE0-

D3(BJ)/def2-QZVPP//BP86/def2-SV(P); COSMO-RS (1-phenylethanol).

determining step was found to be signiﬁcantly higher than the

R−I enables the formation of the Ni(I) complex N and the

393

⧧

−1

ones of the previous systems (ΔG = 156.1 kJ mol ), and the

strong dependence on the length of the ligand’s backbone

observed experimentally was also probed and similarly observed

computationally (see Supporting Information ).

phenylethyl radical (ΔG = 22.8 kJ mol ) that was similarly

formed in the phosphine pathway (cf. Scheme 6). Subsequent

barrierless radical recombination is assumed to instantly occur

•

after the formation of R , leading to the penta-coordinate

−1

Furthermore, the DFT calculations were also employed to

study the reactivity of the system when no phosphine is present

in the reaction mixture (Scheme 6). The starting points on what

type of activated complex could be formed in the reaction are

structure O (ΔG = 50.1 kJ mol ; for a discussion on

competing recombination pathways, see Supporting Informa-

tion ). CO insertion is only subject to a small energy barrier (TS-

⧧

−1

P; ΔG = 61.2 kJ mol ) and will lead to the tetra-coordinate

4,28

−

393

−1

provided by two previous publications:

(1) [Ni(CO) I] Li

acyl nickel complex Q (ΔG = −9.0 kJ mol ). After CO

addition, a slightly endergonic penta-coordinate complex is

formed (R; ΔG = 9.9 kJ mol ) that can undergo reductive

elimination (TS-S), which is also exhibiting a relatively low

activation barrier (ΔG = 52.0 kJ mol ; compared to 106.8 kJ

mol for the reductive elimination in the neutral regime; cf.

−

(

L) is less stable than Ni((CO) L (A), and (2) [Ni(CO) I] Li

−1

is more stable than Ni(CO) (K). Therefore, the Gibbs free

energy of complex L must be between these energies, which is

−1

⧧

−1

conﬁrmed by our computations (A: ΔG = 0.0 kJ mol ; L:

−1

298

−1

ΔG = 36.4 kJ mol ; K: ΔG = 85.6 kJ mol ). Interestingly,

the stability of Ni(CO) (TBP) (or Ni(CO) TBP) signiﬁcantly

Scheme 6) and can readily take place to form the R(CO)I

−1

diﬀers at the conditions employed in this study (pCO = 50 bar, T

120 °C). When accounting for the conditions in the

computations, the stability of the anionic complex surpasses

adduct of [Ni(CO) I]-Li (T; ΔG

= 41.5 kJ mol ).

Dissociation of R(CO)I leads to the regeneration of complex

−1

L (ΔG = −24.4 kJ mol ).

−1

the stability of Ni(CO) (TBP) (A: ΔG = 0.0 kJ mol ; L:

Further quantum-chemical calculations were employed to

investigate the potential involvement of the phosphine ligand in

the anionic pathways to rule out the involvement of [Ni-

298

ΔG393 = −7.1 kJ mol−1; K: ΔG = 42.0 kJ mol−1). The

anionic species is able to carbonylate RI in a similar fashion to

the phosphine-coordinated complex (vide supra). The calcu-

lations on the anionic reaction mechanism were limited to the

−

(CO) (TBP) ] Li or related species (see Supporting Informa-

tion ).

anionic complex, and the Li cation was assumed to be equally

The quantum-chemical investigation supports our assump-

tion that two diﬀerent Ni(I)-based pathways operate at diﬀerent

reaction conditions. (1) Phosphine mechanism: Ni(CO) L or

solvated by 1-phenylethanol and methyl ethyl ketone for all

investigated systems. Similar to the mechanism involving

phosphine-coordinated complexes (vide supra), the anionic

reaction mechanism starts with the formation of the

Ni(CO) L, which are present in equilibrium, can catalyze the

carbonylation via a catalytic cycle in which one phosphine is

coordinated to the nickel center. The mechanism proceeds via

Ni(I) and radical generation, which will subsequently

•

393

corresponding RI adduct, [Ni(CO)2I RI]-Li (M; ΔG

−

3.8 kJ mol ). A (presumably) barrierless homolytic cleavage of

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Scheme 6. Catalytic Cycle for the Carbonylation of R−I (Iodoethylbenzene) without Phosphine

393

−1

The reaction occurs via a “stepwise” oxidative addition at the Ni(I) complex in an anionic regime. ΔG in kJ mol ; RI-PBE0-D3(BJ)/def2-

QZVPP//BP86/def2-SV(P); COSMO-RS (1-phenylethanol).

recombine. Therefore, no “classical” oxidative addition tran-

sition state is involved, which were all computed to be very

More complex and industrially relevant molecules such as

ibuprofen and naproxene were obtained in 52 and 26% yields,

respectively (entries 2 and 3). Substrates with electron-donating

or electron-withdrawing substituents in para positions at the

aromatic ring did not show much inﬂuence on the ﬁnal outcome,

aﬀording similar moderate yields in all cases (entries 2, 4, and 5).

More sterically crowded and cyclic substrates such as 1-

phenylpropanol and 1-indanol yielded 43 and 19% yields,

respectively (entries 6 and 7).

Aliphatic alcohols were also converted to the corresponding

carboxylic acids in good yields under modiﬁed reaction

conditions (Table 5). In this case, higher temperatures were

needed to facilitate the oxidative addition step due to the higher

unfavorable in energy. (2) Anionic mechanism: Based on the

−

previously reported anionic [Ni(CO) I] species, the system

can catalyze the carbonylation by a similar Ni(I) pathway. The

anionic pathway was determined to be signiﬁcantly more

⧧

−1

393

feasible (phosphine: ΔG = 131.7 kJ mol ; anionic: ΔG

−1

0.9 kJ mol ) and will be preferred as long as suﬃcient amounts

of LiI are present in solution. Interestingly, the highest energetic

structure in the anionic mechanistic scenario is the R−I adduct

M, which consequently undergoes homolytic cleavage, while the

CO insertion transition state (TS-E) presents the activation

barrier in the phosphine-based pathway.

Substrate Scope. To evaluate the potential of this nickel-

catalyzed carbonylation, an investigation on the substrate scope

was performed. We decided to use a combination of TBP and LiI

as the results from Table 3 indicate that this combination

performed better for the carbonylation of 1-phenylthanol. We

bond strength of the C_a_l_y_k_l−I bonds compared to the C

−I

benzyl

−

−1

bonds (233.5 ± 6.3 kJ mol vs 187.8 ± 4.8 kJ mol for

iodoethane and benzyl iodide, respectively), as at 120 °C,

mixtures of unreacted alcohol and alkyl iodide were obtained.

When 2-phenylethanol and cyclic alcohols were used as

substrates in the absence of water, carboxylic acids and ester

mixtures were obtained. Water was added to hydrolyze the

formed ester intermediates. The equivalents of water needed

were substrate-dependent and not general for all substrates.

Cyclic alcohols were carbonylated to the corresponding

carboxylic acids in almost quantitative yields (entries 1 and 2).

Linear alcohols aﬀorded the ﬁnal products in moderate to

good yields (entries 3−6). The increased temperatures needed

in these cases resulted in a loss of selectivity, and mixtures of

branched and linear products were obtained. The higher ratio of

also decided to use TBP and LiI instead of PPh or NaI in order

to avoid solubility issues that may turn in lower activity. The

substrate scope was performed using NiI at 4 mol % loading, 8

mol % of TBP, 20 mol % of LiI, 50 bar of CO, at 120 °C for 20 h.

A range of benzylic alcohols were carbonylated to the

corresponding carboxylic acids in moderate to good yields

(Table 4). In all cases, the reaction was selective toward the

branched product. The nonsubstituted 1-phenylethanol and

benzyl alcohol were carbonylated to the corresponding

carboxylic acids in almost quantitative yields (entries 1 and 8).

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Table 4. Screening of Benzylic Alcohols

Table 5. Screening of Aliphatic Alcohols

A mixture of the corresponding alcohol (8.3 mmol), NiI (4 mol %),

TBP (8 mol %), LiI (20 mol %), and water (1 or 2 equiv when

needed; see Experimental Section) in MEK (8 M) was pressurized

with CO at 50 bar and stirred at 200 °C for 20 h. Isolated yields.

H O (2 equiv). H O (1 equiv). No water added. Combined yield

are working in concert when phosphine and LiI are present in

the reaction media. In the absence of halide salt, both ligand and

HI are needed and NiI is reduced to a Ni(0)-carbonyl species at

the start of the reaction. In the presence of phosphine and halide

salts, Ni(0)-carbonyl-phosphine and Ni(0) halogenotricarbonyl

anion species are formed in the reaction media. The HI

generated in the reduction reacts with the alcohol via an S_N1

mechanism to generate alkyl iodide as the reactive intermediate,

which reacts with the Ni(0)-carbonyl species via two one-

electron steps, involving alkyl radicals and halonickel(I)

metalloradical species. Reductive elimination regenerates the

Ni(0) species and releases an acyl halide, which is readily

hydrolyzed to give the corresponding carboxylic acid.

A mixture of the corresponding alcohol (8.3 mmol), NiI (4 mol %),

TBP (8 mol %), and LiI (20 mol %) in MEK (8 M) was pressurized

with CO at 50 bar and stirred at 120 °C for 20 h. Isolated yields.

the linear isomer observed for the carbonylation of 2-

phenylethanol further indicates that the intermediary formed

radical reacts very fast with the Ni(I) complex, as otherwise a

higher degree of isomerization to the more stable benzylic

radical should be observed to deliver more of the branched

product (entry 3).

EXPERIMENTAL SECTION

■

Caution: Carbon monoxide (CO) is an extremely poisonous substance.

Proper safety precautions need to be applied and appropriate safety

equipment MUST be used. Reactions must be carried out in a well-

ventilated fume hood equipped with a CO detector. The CO bottle was

stored outside the laboratory and was permanently connected to a gas

distribution system equipped with magnetic safety valves. The valves

were connected to CO sensors inside the laboratory. Additionally, all of

the reactions involving the use of CO were carried out using a portable

CO detector and personal protection equipment.

CONCLUSIONS

■

We have studied the direct carbonylation of alcohols using a

nickel catalyst under low pressures. The system represents a

straightforward methodology for the preparation of carboxylic

acids and provides highly selective access to branched products.

Only catalytic concentrations of iodide are necessary, making

General Experimental Details. Unless otherwise noted, all

reactions and manipulations were carried out under an Ar atmosphere

using standard Schlenk and high-vacuum-line techniques or in a Braun

inert-atmosphere glovebox (Ar) at ambient temperature. Organic

solvents were purchased dry from Aldrich and degassed prior to use by

the system less corrosive than the currently used Pd-PPh -HCl

process for the synthesis of ibuprofen, where an excess of 20%

HCl is needed. The experimental and quantum-chemical

investigation indicates that two diﬀerent reaction mechanisms

Organometallics XXXX, XXX, XXX−XXX

Organometallics

The Supporting Information is available free of charge at

Article

bubbling argon for at least 30 min. Commercially available chemicals

were purchased from Aldrich, ABCR, or TCI and were used as received

unless otherwise stated. NMR spectra were recorded using a Bruker 200

instrument at CaRLa or Bruker AVANCE III 300, Bruker AVANCE III

C D ): δ 7.66 (dsept, J = 490 Hz, ³J_H_H= 5.6 Hz, 1H), 2.57−2.47 (m,

6H, P−CH −), 1.57−1.48 (m, 6H, −CH −), 1.43−1.33 (m, 6H,

−CH −), 0.91 (t, J = 7.3 Hz, 9H, −CH ). C{ H} NMR (151 MHz,

C D ): δ 25.1 (d, J = 4.7 Hz), 24.1 (d, J = 15.5 Hz), 17.6 (d, J = 46.8

00, Bruker AVANCE III 500, and Bruker AVANCE III 600

Hz), 13.8 (s). P{ H} NMR (243 MHz, C D ): δ 8.83 (s). Anal. Calcd

6 6

spectrometers at the Institute of Organic Chemistry of Heidelberg

University. H and C chemical shifts δ are reported in parts per

For C H PI (330.10 g/mol): C, 43.65; H, 8.55. Found: C, 43.68; H,

8.70.

million relative to either the residual solvent or tetramethylsilane

(

Typical Procedure for the Carbonylation of 1-Phenylethanol

(HEL CAT7 Screening Experiments). The corresponding nickel

source, ligand, and/or additive were placed in a 10 mL microwave glass

vial under Ar atmosphere (glovebox). 1-Phenylethanol (8.28 mmol)

and methyl ethyl ketone (8 M) were added, and the vials were closed

with a septum cap ﬁtted with a needle. The vials were placed in a HEL

CAT7 autoclave which was ﬂushed with carbon monoxide (3 × 5 bar)

and then pressurized with carbon monoxide to 50 atm and heated to

TMS). ³¹P chemical shifts are reported relative to an external standard

of phosphoric acid 30% in D O (0.0 ppm). The multiplicities are

reported as s = singlet, b = broad, d = doublet, t = triplet, q = quartet,

sept = septet, and m = multiplet. GC analyses were performed on an

Agilent Technologies 6890N gas chromatography system equipped

with an FID detector and an Agilent Technologies DB-5 (5%

phenyl)methylpolysiloxane, capillary column (30 m × 0.320 mm ×

20 °C for 20 h. The autoclave was cooled to room temperature, and

.25 μm; He ﬂow 1.0 mL/min; program initial 80 °C, ramp 15 °C/min,

00 °C for 5 min). GC-MS analyses were performed on an Agilent

the remaining pressure was released in a well-ventilated hood. The

crude mixtures were diluted with DCM (2 mL). N-Methyl-N-

Technologies 6890N gas chromatography system coupled with an

Agilent Technologies 5975B mass spectrometer and equipped with an

Agilent Technologies HP-5MS capillary column (30 m × 0.250 mm/

(

trimethylsilyl)triﬂuoroacetamide (0.2 mL), DCM (1.5 mL), and the

reaction mixture (0.1 mL) were added to a GC vial. The vial was heated

on a steel plate to 70 °C for 1 h. The yield and selectivity were

determined by GC-FID analysis using anisole as an internal standard.

Typical Procedure for the Carbonylation of Benzylic

Alcohols (Procedure A). A 15 mL self-made stainless steel autoclave

from Swagelok parts ﬁtted with a Teﬂon insert and a magnetic stirring

.25 μm). Elemental analyses were conducted in house. X-band EPR

spectra were recorded on a Bruker Elexsys E500 instrument equipped

with a ER 4112HV-CF58nc in-cavity cryogen free VT system in

perpendicular mode at 9.63 GHZ at 203 K.

Two types of autoclaves were used for the experiments: a HEL CAT7

autoclave (volume: 7 × 10 mL, internal stirring bar 500 rpm, material:

glass (vials) and stainless steel) and self-made autoclaves from

Swagelok parts (volume 15 mL, internal stirring bar 500 rpm, material:

Teﬂon (insert) and stainless steel). Pictures of the autoclaves are shown

in the Supporting Information .

bar was charged under Ar atmosphere with NiI (4 mol %), TBP (8 mol

), LiI (20 mol %), the corresponding alcohol (8.28 mmol), and

methyl ethyl ketone (8 M). The autoclave was ﬂushed with carbon

monoxide (3 × 5 bar) and then pressurized with carbon monoxide to 50

atm and heated to 120 °C in an oil bath for 20 h. The autoclave was

cooled to room temperature, and the remaining pressure was released in

a well-ventilated hood. Isolated products were obtained by ﬂash

chromatography (Biotage Isolera Prime) using petroleum ether−ethyl

acetate mixtures. Products were identiﬁed according to spectroscopic

data of the commercially available compounds.

Computational Details. All geometry optimizations were carried

5−57

out at the BP86/def2-SV(P)

level of theory with relativistic

corrected eﬀective core potentials for iodine. Stationary points were

veriﬁed via analysis of the vibrational frequencies at the level of

geometry optimization. Final electronic energies were obtained by

single-point calculations at the PBE0-D3(BJ)/def2-QZVPP

of theory employing Grimme’s D3 dispersion correction incorporat-

ing Becke-Johnson damping. All quantum-chemical calculations were

carried out using the TURBOMOLE program

the resolution of identity (RI) approximation

57,59,60

Typical Procedure for the Carbonylation of Aliphatic

Alcohols (Procedure B). A 15 mL self-made stainless steel autoclave

from Swagelok parts ﬁtted with a Teﬂon insert and a magnetic stirring

level

3−65

bar was charged under Ar atmosphere (glovebox) with NiI (4 mol %),

(version 7.3) with

6−69

TBP (8 mol %), LiI (20 mol %), the corresponding alcohol (8.28

mmol), and methyl ethyl ketone (8 M). Outside the glovebox, the

corresponding equivalents, if needed, of degassed water were added.

The autoclave was ﬂushed with carbon monoxide (3 × 5 bar) and then

pressurized with carbon monoxide to 50 atm and heated to 200 °C in an

oil bath for 20 h. The autoclave was cooled to room temperature, and

the remaining pressure was released in a well-ventilated hood. Isolated

products were obtained by ﬂash chromatography (Biotage Isolera

Prime) using petroleum ether−ethyl acetate mixtures. Products were

identiﬁed according to spectroscopic data of the commercially available

compounds.

and the correspond-

70,71

ing auxiliary basis sets

implemented in the program. Zero-point

vibrational energies and thermodynamic corrections were obtained at

the level of geometry optimization and scaled to the given reaction

temperatures (120 °C). For all species, the thermodynamic reference

concentration was set to x = 0.01, except for CO (p = 50 bar). Solvent

corrections to Gibbs free energies in 1-phenylethanol were calculated

for all species, except CO and CO , with the conductor-like screen

72,73

model for real solvents (COSMO-RS)

carried out with the

(version 18.0.0; revision 4360; parameters

BP_TZVP_18.ctd). All energies discussed are Gibbs free energies (G)

74,75

COSMOtherm program

−

in kJ mol . Connectivities between minima and transition states

ASSOCIATED CONTENT

sı Supporting Information

implied in ﬁgures and schemes were validated by intrinsic reaction

■

coordinates following calculations₇. Pictures of molecular structures

were generated with the CYLview program.

https://pubs.acs.org/doi/10.1021/acs.organomet.0c00082.

A multitude of geometrical isomers were analyzed for all complexes.

Due to the vast number of structures, we have decided to limit the

structures reported herein to the one lowest in energy for each species.

Our attempts to locate transitions states for the homolytic cleavage of

R−I or the radical recombination of the metal-centered radicals and R•

were unsuccessful. Neither closed-shell nor open-shell methods

Cartesien coordinates of the calculated structures (XYZ)

Details regarding screening experiments, pictures of the

autoclaves used, computational details, detailed analyses

and spectra of all products (PDF )

(broken symmetry) nor potential energy surface scans revealed any

signiﬁcant barrier. We therefore propose these steps to be of barrierless

nature; however, we cannot exclude that this is an artifact of the single-

reference methods.

AUTHOR INFORMATION

■

Synthesis of TBP·HI. In a Schlenk ﬂask, a 57% aqueous solution of

HI (1.3 mL, 9.8 mmol) was added dropwise to TBP (2 mL, 8 mmol)

under argon at room temperature. The mixture was stirred for 18 h and

then extracted using dichloromethane/water mixtures. The organic

layers were combined, and the solvent was removed under reduced

Corresponding Author

Thomas Schaub − Catalysis Research Laboratory (CaRLa),

9120 Heidelberg, Germany; BASF SE, Organic Synthesis,

67056 Ludwigshafen, Germany; orcid.org/0000-0003-

pressure to aﬀord a white solid. Yield: 2.3 g, 87%. H NMR (600 MHz,

2332-0376 ; Email: thomas.schaub@basf.com

Organometallics XXXX, XXX, XXX−XXX

Organometallics

(15) Rizkalla, N. Acetic acid production via low-pressure, Nickel-

Article

Authors

(11) Le Berre, C.; Serp, P.; Kalck, P.; Torrence, G. P. Acetic acid.

Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH: Weinheim,

Germany, 2014.

Sara Sabater − Catalysis Research Laboratory (CaRLa), 69120

Heidelberg, Germany

(12) Eastham, G. R.; Richards, P. I. A process for the carbonylation of

Maximilian Menche − Catalysis Research Laboratory (CaRLa),

ethylenically unsaturated compound and a catalytic system. WO Patent

Appl. WO 2009/010782 A1, 2009.

9120 Heidelberg, Germany; BASF SE, Quantum Chemistry,

7056 Ludwigshafen, Germany

(13) Samel, U.; Kohler, W.; Gamer, A. O.; Keuser, U.; Yang, S.; Jin, Y.;

Tamal Ghosh − Catalysis Research Laboratory (CaRLa), 69120

Heidelberg, Germany

Saskia Krieg − Anorganisch-Chemisches Institut, Heidelberg

University, 69120 Heidelberg, Germany

Lin, M.; Wang, Z. Propionic acid and derivatives. Ullmann’s

Encyclopedia of Industrial chemistry; Wiley-VCH: Weinheim, Germany,

014.

(14) Elango, V.; Murphy, M. A.; Smith, B. L.; Mott, G. N.; Moss, G. L.

Process for producing ibuprofen. EP Patent Appl. EP0284310A1, 1988.

catalyzed methanol carbonylation. ACS Sym. Ser. 1987, 328, 61−76.

Katharina S. L. Ruck − BASF SE, Organic Synthesis, 67056

Ludwigshafen, Germany

Rocco Paciello − BASF SE, Organic Synthesis, 67056

Ludwigshafen, Germany

Ansgar Schafer − BASF SE, Quantum Chemistry, 67056

Ludwigshafen, Germany

Peter Comba − Anorganisch-Chemisches Institut, Heidelberg

University, 69120 Heidelberg, Germany; orcid.org/0000-

16) Kelkar, A. A.; Ubale, R. S.; Deshpande, R. M.; Chaudhari, R. V.

Carbonylation of methanol using Nickel-complex catalyst - A kinetic-

study. J. Catal. 1995, 156, 290−294.

(17) Ubale, R. S.; Kelkar, A. A.; Chaudhari, R. V. Carbonylation of

ethanol using Ni-isoquinoline complex catalyst: Activity and selectivity

studies. J. Mol. Catal. A: Chem. 1997, 118, 9−19.

001-7796-3532

A. Stephen K. Hashmi − Catalysis Research Laboratory

CaRLa), 69120 Heidelberg, Germany; Organisch-Chemisches

(18) Kelka, A. A.; Ubale, R. S.; Chaudhari, R. V. Promoting effects of

metal iodides in Ni(Isoq)4Cl2-catalyzed carbonylation of ethanol and

n-propanol. J. Catal. 1992, 136, 605−608.

(19) Chaudhari, R. V.; Seayad, J.; Seayad, A. Process for the

preparation of 2-aryl propionic acids. U.S. Patent Appl. 6,660,883 B1,

(

Institut, Heidelberg University, 69120 Heidelberg, Germany;

orcid.org/0000-0002-6720-8602

(

003.

20) Tanaka, K.; Shima, Y. Process for production of alpha-(4-

isobutylphenyl)propionic acid. EP Patent Appl. EP0361021A1, 1990.

21) Fell, B. Hydrocarboxylation reaction of unsaturated-compounds

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.organomet.0c00082

with Nickel carbonyl-complex and cobalt carbonyl-complex catalysts.

Author Contributions

S.S. and M.M. contributed equally.

Fette, Seifen, Anstrichm. 1974, 76, 193−196.

(22) Mitsubishi Gas Chemical Company Inc. Process for production

Notes

of alpha-(4-isobutylphenyl)propionic acid. EP Patent Appl.

EP0361021B1, 1990.

The authors declare no competing ﬁnancial interest.

(

23) Sanz, R.; Miguel, D.; Martinez, A.; Alvarez-Gutierrez, J. M.;

ACKNOWLEDGMENTS

Rodriguez, F. Bronsted acid-catalyzed benzylation of 1,3-dicarbonyl

■

derivatives. Org. Lett. 2007 , 9 , 2027 −2030.

The authors thank the analytical departments of the Institute for

Organic Chemistry for excellent support. CaRLa (Catalysis

Research Laboratory) is coﬁnanced by Heidelberg University

and BASF SE.

(24) Rizkalla, N. Acetic-Acid Production Via Low-Pressure, Nickel-

Catalyzed Methanol Carbonylation. Acs Sym. Ser. 1987 , 328 , 61 −76.

(25) Giannoccaro, P.; Vasapollo, G. Carbon-Monoxide Reaction

Towards Ni(Ii) Tetracoordinated Phosphine Complexes In Aqueous

Alcoholic Solution. Inorg. Chim. Acta 1983, 72, 51−55.

REFERENCES

(26) Matsumoto, S.; Naito, M.; Oseki, T.; Akazome, M.; Otani, Y.

Selective reaction of benzyl alcohols with HI gas: Iodination, reduction,

and Indane ring formations. Tetrahedron 2017, 73 , 7254 −7259.

■

(

1) Bertleﬀ, W.; Roeper, M.; Sava, X. Carbonylation. Ullmann’s

Encyclopedia of Industrial Chemistry; Wiley-VCH: Weinheim, Germany,

(27) Kelka, A. A.; Ubale, R. S.; Chaudhari, R. V. Promoting Effects Of

(

012.

Metal Iodides In Ni(Isoq)4Cl2-Catalyzed Carbonylation Of Ethanol

2) Kubitschke, J.; Lange, H.; Strutz, H. Carboxylic acids aliphatic.

And Normal-Propanol. J. Catal. 1992 , 136 , 605 −608.

Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH: Weinheim,

Germany, 2014.

(28) Cassar, L.; Foa, M. Halogentricarbonylnickel anions. Inorg. Nucl.

Chem. Lett. 1970, 6, 291−294.

(3) Wu, X. F.; Fang, X. J.; Wu, L. P.; Jackstell, R.; Neumann, H.; Beller,

(29) Cassar, L.; Foa, M. Nickel-catalyzed carbonylation of aromatic

M. Transition-metal-catalyzed carbonylation reactions of olefins and

halides at atmospheric-pressure of carbon-monoxide. J. Organomet.

Chem. 1973, 51, 381−393.

alkynes: A personal account. Acc. Chem. Res. 2014 , 47 , 1041 −1053.

(4) Brennfuhrer, A.; Neumann, H.; Beller, M. Palladium-Catalyzed

(30) Stille, J. K.; Cowell, A. B. Oxidative addition of benzyl halides to

Carbonylation Reactions of Aryl Halides and Related Compounds.

Angew. Chem., Int. Ed. 2009 , 48 , 4114 −4133.

tetrakis(triphenylphosphine)Nickel(0). J. Organomet. Chem. 1977, 124,

253−261.

(31) Perez Garcia, P. M.; Di Franco, T.; Orsino, A.; Ren, P.; Hu, X. L.

Nickel-catalyzed diastereoselective alkyl-alkyl Kumada coupling

reactions. Org. Lett. 2012, 14, 4286−4289.

(32) Kehoe, R.; Mahadevan, M.; Manzoor, A.; McMurray, G.;

Wienefeld, P.; Baird, M. C.; Budzelaar, P. H. M. Reactions of the Ni(0)

compound Ni(PPh3)(4) with unactivated alkyl halides: oxidative

addition reactions involving radical processes and Nickel(I) inter-

mediates. Organometallics 2018, 37, 2450−2467.

(33) Manzoor, A.; Wienefeld, P.; Baird, M. C.; Budzelaar, P. H. M.

Catalysis of cross-coupling and homocoupling reactions of aryl halides

utilizing Ni(0), Ni(I), and Ni(II) precursors; Ni(0) compounds as the

probable catalytic species but Ni(I) compounds as intermediates and

products. Organometallics 2017, 36, 3508−3519.

(5) Barnard, C. F. J. Palladium-catalyzed carbonylation-A reaction

come of age. Organometallics 2008, 27, 5402−5422.

6) Brennfuhrer, A.; Neumann, H.; Beller, M. Palladium-catalyzed

carbonylation reactions of alkenes and alkynes. ChemCatChem 2009, 1,

8−41.

7) Paulik, F. E.; Roth, J. F. Novel catalysts for low-pressure

carbonylation of methanol to acetic acid. Chem. Commun. 1968, 1578a.

8) Zoeller, J. R.; Cloyd, J. D.; Lafferty, N. L.; Nicely, V. A.;

Polichnowski, S. W.; Cook, S. L. Rhodium-catalyzed carbonylation of

methyl acetate. Adv. Chem. Ser. 1992, 230, 377−394.

(9) Garland, C. S.; Giles, M. F.; Sunley, J. G. Process and catalyst for

the production of acetic acid. EP Patent Appl. EP0643034B1, 1995.

10) Haynes, A. Catalytic methanol carbonylation. Advances in

Catalysis; Elsevier Inc.: Amsterdam, 2010; Vol. 53, pp 1−45.

(

Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

34) Funes-Ardoiz, I.; Nelson, D. J.; Maseras, F. Halide abstraction

(55) Perdew, J. P. Density-functional approximation for the

correlation energy of the inhomogeneous electron gas. Phys. Rev. B:

Condens. Matter Mater. Phys. 1986 , 33 , 8822 −8824.

competes with oxidative addition in the reactions of aryl halides with

Ni(PMenPh(3-n))(4). Chem. - Eur. J. 2017 , 23 , 16728 −16733.

catalyzed reductive carbonylation utilizing ethyl chloroformate as the

carbonyl source. Angew. Chem., Int. Ed. 2019, 58, 7454−7458.

35) Shi, R. Y.; Hu, X. L. From alkyl halides to ketones: Nickel-

(56) Becke, A. D. Density-functional exchange-energy approximation

with correct asymptotic behavior. Phys. Rev. A: At., Mol., Opt. Phys.

1988, 38, 3098−3100.

36) Breitenfeld, J.; Ruiz, J.; Wodrich, M. D.; Hu, X. L. Bimetallic

(57) Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence,

triple zeta valence and quadruple zeta valence quality for H to Rn:

Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7,

3297−3305.

oxidative addition involving radical intermediates in Nickel-catalyzed

alkyl-alkyl Kumada coupling reactions. J. Am. Chem. Soc. 2013, 135,

37) Diccianni, J. B.; Katigbak, J.; Hu, C. H.; Diao, T. N. Mechanistic

2004−12012.

(58) Andrae, D.; Haußermann, U.; Dolg, M.; Stoll, H.; Preuß, H.

Characterization of (Xantphos)Ni(I)-Mediated Alkyl Bromide Activa-

tion: Oxidative Addition, Electron Transfer, or Halogen-Atom

Abstraction. J. Am. Chem. Soc. 2019 , 141 , 1788 −1796.

Energy-adjusted ab initio pseudopotentials for the second and third row

transition elements. Theoretica chimica acta 1990 , 77 , 123 −141.

(59) Ernzerhof, M.; Scuseria, G. E. Assessment of the Perdew −

Burke −Ernzerhof exchange-correlation functional. J. Chem. Phys. 1999,

110, 5029−5036.

(38) Tsou, T. T.; Kochi, J. K. Mechanism of oxidative addition -

reaction of Nickel(0) complexes with aromatic halides. J. Am. Chem.

Soc. 1979, 101, 6319−6332.

(60) Weigend, F.; Furche, F.; Ahlrichs, R. Gaussian basis sets of

quadruple zeta valence quality for atoms H −Kr. J. Chem. Phys. 2003,

(39) Liu, Y.; Cornella, J.; Martin, R. Ni-Catalyzed Carboxylation of

61) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and

19, 12753−12762.

Unactivated Primary Alkyl Bromides and Sulfonates with CO2. J. Am.

Chem. Soc. 2014, 136, 11212−11215.

accurate ab initio parametrization of density functional dispersion

correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010,

(40) Nilges, M. J.; Barefield, E. K.; Belford, R. L.; Davis, P. H.

Electronic-structure of 3-coordinate Nickel (I) - electron-paramagnetic

resonance of Nickel-doped halobis(triphenylphosphine)Copper(I). J.

Am. Chem. Soc. 1977, 99, 755−760.

62) Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the damping

32, 154104−154119.

function in dispersion corrected density functional theory. J. Comput.

(41) Wertz, J. E.; Bolton, J. R. Electron Spin Resonance: Elementary

Chem. 2011, 32, 1456−1465.

Theory and Practical Applications; Chapman and Hall: New York, 1986.

42) Bottoni, A.; Miscione, G. P.; Novoa, J. J.; Prat-Resina, X. DFT

(63) TURBOMOLE, version 7.3, a development of University of

Computational Study of the Mechanism of Allyl Halides Carbonylation

Karlsruhe and Forschungszentrum Karlsruhe GmbH; TURBOMOLE

GmbH, 2018, available from http://www.turbomole.com.

Catalyzed by Nickel Tetracarbonyl. J. Am. Chem. Soc. 2003, 125,

64) Ahlrichs, R.; Bar, M.; Haser, M.; Horn, H.; Kolmel, C. Electronic

̈ ̈ ̈

structure calculations on workstation computers: The program system

0412−10419.

43) Bottoni, A.; Miscione, G. P.; Carvajal, M. A.; Novoa, J. J. The

turbomole. Chem. Phys. Lett. 1989, 162 , 165 −169.

mechanism of transition metal catalyzed carbonylation of allyl halides:

66) Vahtras, O.; Almlof, J.; Feyereisen, M. W. Integral approx-

65) Treutler, O.; Ahlrichs, R. Efficient molecular numerical

A theoretical investigation. J. Organomet. Chem. 2006 , 691 , 4498 −4507.

integration schemes. J. Chem. Phys. 1995, 102, 346−354.

imations for LCAO-SCF calculations. Chem. Phys. Lett. 1993, 213,

14−518.

67) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R.

(44) Breitenfeld, J.; Wodrich, M. D.; Hu, X. Bimetallic Oxidative

Addition in Nickel-Catalyzed Alkyl −Aryl Kumada Coupling Reactions.

Organometallics 2014, 33, 5708−5715.

(45) Wotal, A. C.; Ribson, R. D.; Weix, D. J. Stoichiometric Reactions

of Acylnickel(II) Complexes with Electrophiles and the Catalytic

Auxiliary basis sets to approximate Coulomb potentials. Chem. Phys.

Lett. 1995, 240, 283−290.

Synthesis of Ketones. Organometallics 2014, 33 , 5874 −5881.

(46) Zhao, C.; Jia, X.; Wang, X.; Gong, H. Ni-Catalyzed Reductive

(68) Deglmann, P.; May, K.; Furche, F.; Ahlrichs, R. Nuclear second

Coupling of Alkyl Acids with Unactivated Tertiary Alkyl and Glycosyl

analytical derivative calculations using auxiliary basis set expansions.

Halides. J. Am. Chem. Soc. 2014 , 136 , 17645 −17651.

Chem. Phys. Lett. 2004 , 384 , 103 −107.

(47) Weix, D. J. Methods and Mechanisms for Cross-Electrophile

(69) Weigend, F. Accurate Coulomb-fitting basis sets for H to Rn.

Coupling of Csp2 Halides with Alkyl Electrophiles. Acc. Chem. Res.

015, 48, 1767−1775.

48) Andersen, T. L.; Donslund, A. S.; Neumann, K. T.; Skrydstrup, T.

Phys. Chem. Chem. Phys. 2006, 8, 1057−1065.

70) Weigend, F.; Haser, M.; Patzelt, H.; Ahlrichs, R. RI-MP2:

optimized auxiliary basis sets and demonstration of efficiency. Chem.

Phys. Lett. 1998, 294, 143−152.

Carbonylative Coupling of Alkyl Zinc Reagents with Benzyl Bromides

Catalyzed by a Nickel/NN2Pincer Ligand Complex. Angew. Chem., Int.

Ed. 2018, 57, 800−804.

71) Hellweg, A.; Hattig, C.; Hofener, S.; Klopper, W. Optimized

̈ ̈

accurate auxiliary basis sets for RI-MP2 and RI-CC2 calculations for the

(49) Donslund, A. S.; Neumann, K. T.; Corneliussen, N. P.; Grove, E.

atoms Rb to Rn. Theor. Chem. Acc. 2007 , 117 , 587 −597.

K.; Herbstritt, D.; Daasbjerg, K.; Skrydstrup, T. Access to β -Ketonitriles

through Nickel-Catalyzed Carbonylative Coupling of α -Bromonitriles

with Alkylzinc Reagents. Chem. - Eur. J. 2019 , 25 , 9856 −9860.

(72) Klamt, A. Conductor-like Screening Model for Real Solvents: A

New Approach to the Quantitative Calculation of Solvation

Phenomena. J. Phys. Chem. 1995, 99, 2224−2235.

(50) Zhao, S.; Mankad, N. P. Metal-catalysed radical carbonylation

74) Geilen, F. M. A.; Engendahl, B.; Ho

73) Klamt, A.; Jonas, V.; Burger, T.; Lohrenz, J. C. W. Refinement

reactions. Catal. Sci. Technol. 2019, 9, 3603−3613.

and Parametrization of COSMO-RS. J. Phys. Chem. A 1998, 102, 5074−

085.

(51) Fukuyama, T.; Nishitani, S.; Inouye, T.; Morimoto, K.; Ryu, I.

(

Effective acceleration of atom transfer carbonylation of alkyl iodides by

metal complexes. Application to the synthesis of the hinokinin

precursor and dihydrocapsaicin. Org. Lett. 2006, 8, 1383−6.

lscher, M.; Klankermayer, J.;

Leitner, W. Selective Homogeneous Hydrogenation of Biogenic

Carboxylic Acids with [Ru(TriPhos)H]+: A Mechanistic Study. J.

Am. Chem. Soc. 2011, 133, 14349 −14358.

(52) Fusano, A.; Sumino, S.; Nishitani, S.; Inouye, T.; Morimoto, K.;

Fukuyama, T.; Ryu, I. Pd/Light-Accelerated Atom-Transfer Carbon-

ylation of Alkyl Iodides: Applications in Multicomponent Coupling

Processes Leading to Functionalized Carboxylic Acid Derivatives.

Chem. - Eur. J. 2012, 18, 9415−9422.

(75) Eckert, F.; Klamt, A. Fast solvent screening via quantum

chemistry: COSMO-RS approach. AIChE J. 2002 , 48 , 369 −385.

76) Fukui, K. The path of chemical reactions - the IRC approach. Acc.

Chem. Res. 1981, 14, 363−368.

77) Legault, C. Y. Cylview 1.0b ; Universite

available from http://www.cylview.org .

(

(53) Goossen, L. J.; Koley, D.; Hermann, H. L.; Thiel, W. Mechanistic

(

de Sherbrooke, 2009,

Pathways for Oxidative Addition of Aryl Halides to Palladium(0)

Complexes: A DFT Study. Organometallics 2005, 24, 2398−2410.

(54) CRC Handbook of Chemistry and Physics, 96th ed.; Haynes, W. H.,

Ed.; CRC Press: Boca Raton, FL, 2016.